Quantum-Dot-Sensitized Solar Cells - ACS Publications - American

Article pubs.acs.org/Langmuir

Quantum-Dot-Sensitized Solar Cells: Understanding Linker Molecules through Theory and Experiment Johannes T. Margraf,†,‡ Andrés Ruland,† Vito Sgobba,† Dirk M. Guldi,*,† and Timothy Clark*,‡ †

Department of Chemistry and Pharmacy and Interdisciplinary Center of Molecular Materials (ICMM), Friedrich-Alexander-University Erlangen-Nuremberg, Egerlandstraße 3, 91058 Erlangen Germany ‡ Department of Chemistry and Pharmacy, Computer-Chemistry-Center and Interdisciplinary Center of Molecular Materials (ICMM), Friedrich-Alexander-University Erlangen-Nuremberg, Nägelsbachstraße 25, 91058 Erlangen, Germany S Supporting Information *

ABSTRACT: We have investigated the role of linker molecules in quantum-dot-sensitized solar cells (QDSSCs) using densityfunctional theory (DFT) and experiments. Linkers not only govern the number of attached QDs but also influence charge separation, recombination, and transport. Understanding their behavior is therefore not straightforward. DFT calculations show that mercaptopropionic acid (MPA) and cysteine (Cys) exhibit characteristic binding configurations on TiO2 surfaces. This information is used to optimize the cell assembly process, yielding Cys-based cells that significantly outperform MPA cells, and reach power conversion efficiencies (PCE) as high as 2.7% under AM 1.5 illumination. Importantly, the structural information from theory also helps understand the cause for this improved performance.

D

ye-sensitized solar cells (DSSCs) have proven to be a promising emerging photovoltaic technology1 but still suffer from high cost and limited lifetimes. QDs are promising alternatives to molecular dyes because of their ease of preparation and their broad and tunable absorption.2−5 QDs also display exciting properties, such as possible multiple exciton generation and efficient cascade resonant energy transfer between stacked QD layers of different diameters, which can potentially enhance photon-to-current efficiencies.6,7 State-of-the-art QDSSCs display PCEs of around 5%, while depleted heterojunction cells, which are QD-multilayer-based solid-state devices, reach 7%.8 Currently, the most-studied QDSSC architecture relies on CdSe-QDs, a ZnS passivation layer, an aqueous polysulfide electrolyte, and a Cu2S counter electrode (Figure 1, top).9,10 This setup no longer gives record efficiencies, but it constitutes an ideal test-bed for studying new paradigms, similarly to P3HT/PCBM in organic photovoltaics and N719/iodide in DSSCs. The two main approaches for sensitizing mesoporous electrodes with QDs are to use presynthesized QDs or to form them in situ at the electrode. The latter approach allows CdSe-QD cells with relatively high PCE (around 4% for purely CdSe-sensitized cells) to be constructed using deposition methods such as chemical bath deposition (CBD) and successive ionic layer adhesion and reaction (SILAR).11 These high efficiencies result from the high quantum-dot loading of the photoelectrodes obtainable with these methods. Unfortunately, in situ approaches allow little control over the diameter and shape of the deposited QDs. Some unique © 2013 American Chemical Society

Figure 1. Top: schematic representation of a QDSSC. Bottom: DFTcalculated adsorption geometries of MPA (a, b) and Cys (c, d), including formal charges.

Received: December 3, 2012 Revised: January 24, 2013 Published: January 24, 2013 2434

dx.doi.org/10.1021/la3047609 | Langmuir 2013, 29, 2434−2438

Langmuir

Article

Figure 2. Top: J−V curves of QDSSCs, assembled at pH 1, 8, 4.5, and 6 (from light to dark). Bottom: AFM images of a bare TiO2 film (left) and Cys-QD sensitized films assembled at pH 8 (center) and 6 (right).

on the TiO2 surface. For Cys, a zwitterionic bidentate structure was found to be most stable (by ∼8 kcal mol−1). The most stable geometries and the corresponding adsorption energies are shown in Figure 1, bottom. The predominance of the zwitterionic structure of Cys on TiOx was further corroborated by FTIR spectroscopy (Figures S1−S3 and Table S1), which reveals NH3+ but no primary amine bands. This correlates well with the large difference in energy between structures c and d, which suggests that the zwitterionic species is formed almost exclusively at thermal equilibrium. Interestingly, Cys is usually absorbed onto TiOx from organic solvents such as toluene, even though it is only poorly soluble in such solvents.14 We decided to use aqueous solutions because the DFT results suggested that the pH of an aqueous Cys solution should have an impact on Cys adsorption. Both Cys and TiOx can be charged in aqueous solutions, depending on the pH. The isoelectric point of Cys (IEP) lies at pH 6,21 and the point of zero charge (PZC) of anatase has an average value of pH 5.9.22 This means that beneath pH 5.9 both components are positively charged, while above pH 6 they are negatively charged. At the IEP, Cys exists in the zwitterionic structure, which is most stable on the surface. The best results for Cys absorption on anatase should therefore be observed close to pH 6. In fact, a similar effect has been reported for the direct adsorption of MPA-capped CdSe QDs to TiOx.12 To test this relationship, we prepared a set of Cys-TiOx electrodes using Cys solutions with different pH. As expected, the short circuit current (Jsc), fill factor (FF), and consequently PCE depend on the pH of the linker solution (Figure 2, top). In particular, the cells in which Cys was deposited at pH 6 performed best with a PCE of 1.90%. Depositing the QDs at pH 1 resulted in a PCE of 1.46%, while at pH 4.5 and 8 the PCEs are 1.57%. We attribute the remarkable performance at

properties of QDs, especially their tunable absorption, can therefore not be used systematically to improve the solar cells. The use of presynthesized QDs, on the other hand, allows exact size control, which facilitates rational solar-cell design. In this case, the QDs must be attached to the electrode via a bifunctional linker molecule. The downside of this approach is that reported PCEs for this type of QDSSC are usually significantly lower than for their in situ analogues (1−2% for CdSe).12 However, a recent publication reports a linker-based CdSe QDSSC (employing MPA) with a PCE of 5.4%, indicating that low performance is a design issue and not a fundamental one.13 We now report work that complements previous studies, which indicated that Cys displays a more efficient charge injection than the conventional linker MPA.14−16 The inferior performance of Cys-based devices led to the suggestion that Cys is an inherently unsuitable linker that causes faster charge recombination. We now show that this is not the case and that the reported lower PCE of Cys was due solely to lower QD loading. In fact, our optimized procedure overcomes this limitation. Notably, Cys is also much cheaper and less toxic than MPA. We performed DFT simulations using a procedure suggested by Troisi.17 Specifically, we placed a linker molecule in the center of a 2 × 2 anatase supercell in order to model the linker monolayers on TiO2. This vacuum slab had an exposed (101) surface and was formed of three TiO2 layers. We assumed that the molecules would bind primarily through the carboxylate group. A possible coordination of the amino group was also considered for Cys. The linkers were placed in different starting geometries, and the geometries were optimized using the PBE functional and a DNP basis set in DMol3.18−20 The calculations revealed several possible adsorbate structures. The most stable for MPA (by ∼2 kcal mol−1) was bidentate bridging, with a proton transferred to an oxygen atom 2435

dx.doi.org/10.1021/la3047609 | Langmuir 2013, 29, 2434−2438

Langmuir

Article

pH 6 to the maximum QD loading under these conditions because the Cys coverage is highest at this pH. This hypothesis was corroborated by atomic force microscopy (AFM) (Figure 2, bottom). We prepared thin TiOx films on glass from a TiCl4 solution and grafted Cys at pH 6 and 8 prior to the QD uptake. New features are discernible in the presence of QDs, whose density increases when going from pH 8 to 6. In particular, heights between 6 and 8 nm were observed (Figures S4−S7). These are consistent with bound QDs, if we consider QD diameters of ∼3.6 nm and the presence of the oleic acid ligand shell. Moreover, the fact that the pH dependence is reproduced on flat TiOx films rules out different penetration of Cys inside the pores of the mesoscopic electrodes as the decisive step. Instead, pH-dependent electrostatic repulsions between TiOx and Cys govern the deposition. Absorption spectra of the photoelectrodes (Figure S8) further confirm this hypothesis. As such, formation of a dense monolayer of linker constitutes the overall bottleneck in obtaining high-efficiency QDSSCs in this approach. Studies on rutile single crystals have shown a significantly higher coverage for MPA-linked films.23 We suspect that the less ideal surface used in our case represents a more realistic model system, since the observed coverage is more in line with estimates for mesoporous films.25 To optimize the cell-assembly process further, the thickness of the TiOx layer was increased by applying two layers of TiOx paste, increasing the concentration of the electrolyte to 2 M and omitting the second TiCl4 treatment. Although the latter has been reported to increase the efficiency of DSSCs by improving the connectivity between TiOx particles,24 we found that this was not the case for QDSSCs. We speculate that the treatment decreases the pore size, lowering the potential for QD adsorption. These simple changes brought a further massive improvement in efficiency to a maximum PCE = 2.7%, as shown in Figure 3. This value is quite high for this type of cell in general, independent of the linker molecule. Comparing this to a similarly optimized MPA cell shows that Cys is indeed superior to MPA (PCE = 1.1%), if the conditions used to form the linker monolayer are chosen correctly. To characterize the effect of the linkers on the photocurrent generation mechanism, we performed electrochemical impedance spectroscopic measurements (EIS). The spectra were recorded at Voc and under AM 1.5 illumination (Figure 4). The Cys cell displays a smaller arc in the mid frequency region of the Nyquist plots, which is attributed to charge recombination.26,27 Therefore, higher recombination resistance is observed for the Cys-sample. The lifetime of injected electrons, which was calculated from the peak frequency in the Bode plot, was only slightly shorter than that for MPA: 28 ms for MPA versus 25 ms for Cys. The electron lifetime in TiOx is limited by the recombination rate. This factor shows very little dependence on the linker in our measurements. This can be explained by the fact that recombination occurs mainly between electrons in the TiOx conduction band and the electrolyte. The surface coverage in QDSSCs of this type has been shown to be well below 20%.25 This means that for both Cys- and MPA-based cells there is a relatively large direct TiOx/electrolyte interfacea trend also found in our AFM investigations. If recombination takes place mainly at this interface, it should naturally not be affected by the linker. Therefore, the proposed fast recombination caused by Cys is not a hindrance for high-performance solar cells.

Figure 3. Top: J−V curves of optimized QDSSCs, assembled from Cys (black) and MPA (red). Bottom: IPCE spectrum of the Cys cell.

We attribute the reported fast charge injection of Cys linked QDs, which is important for achieving high currents, to its zwitterionic structure. The nitrogen atom bears a positive formal charge, which, together with the formal negative charge at the carboxylate group, forms a dipole. This favors movement of electrons from QD to TiOx and hinders it in the reverse direction. For MPA, on the other hand, the TiOx surface is protonated, forming an inverse dipole with the carboxylate. The higher efficiency of Cys cells compared with the MPA cells cannot be due entirely to the dipole moment of the linker. Interestingly, absorption spectra that were taken of the photoelectrodes corroborate that even employment of optimized conditions for Cys adsorption, MPA cells still display higher QD loadings (see Figure S10). AFM images of flat TiOx films are quite revealing (Figure S9), as well as SEM images of mesoporous films (Figures S11−14). No doubt, large amounts of QDs attach to the TiOx surface. Nevertheless, much more intense QD aggregation causes a significant amount of QDs to be electronically decoupled from TiOx. This wellknown phenomenon for MPA28 is not observed for Cys. In conclusion, we have revised the protocol for constructing Cys-based QDSSCs. This allows efficient QDSSCs with the standard cell architecture (i.e., TiOx/linker/QD/ZnS/polysulfide/Cu2S) to be constructed. The combination of DFT simulations, AFM, and EIS data has provided insight into the form and role of the linker layer. Our results suggest that the higher charge-injection efficiency of Cys compared to MPA is due to its zwitterionic structure and that charge recombination 2436

dx.doi.org/10.1021/la3047609 | Langmuir 2013, 29, 2434−2438

Langmuir

Article

We thank Prof. Bernd Meyer for advice on periodic calculations on anatase surfaces and Jenny Malig for help with the AFM.

■

(1) O’Regan, B.; Grätzel, M. A Low-Cost, High-Efficiency Solar Cell Based on Dye-Sensitized Colloidal Titanium Dioxide Films. Nature 1991, 353, 737−740. (2) Mora-Seró, I.; Giménez, S.; Fabregat-Santiago, F.; Gómez, R.; Shen, Q.; Toyoda, T.; Bisquert, J. Recombination in Quantum Dot Sensitized Solar Cells. Acc. Chem. Res. 2009, 42, 1848−1857. (3) Rühle, S.; Shalom, M.; Zaban, A. Quantum-Dot-Sensitized Solar Cells. ChemPhysChem 2010, 11, 2290−2304. (4) Yang, Z.; Chen, C.-Y.; Roy, P.; Chang, H.-T. Quantum DotSensitized Solar Cells Incorporating Nanomaterials. Chem. Commun. 2011, 47, 9561−9571. (5) Kamat, P. V. Quantum Dot Solar Cells. Semiconductor Nanocrystals as Light Harvesters. J. Phys. Chem. C 2008, 112, 18737−18753. (6) Sambur, J. B.; Novet, T.; Parkinson, B. A. Multiple Exciton Collection in a Sensitized Photovoltaic System. Science 2010, 330, 63− 66. (7) Ruland, A.; Schulz-Drost, C.; Sgobba, V.; Guldi, D. M. Enhancing Photocurrent Efficiencies by Resonance Energy Transfer in CdTe Quantum Dot Multilayers: Towards Rainbow Solar Cells. Adv. Mater. 2011, 23, 4573−4577. (8) Ip, A. H.; Thon, S. M.; Hoogland, S.; Voznyy, O.; Zhitomirsky, D.; Debnath, R.; Levina, L.; Rollny, L. R.; Carey, G. H.; Fischer, A.; Kemp, K. W.; Kramer, I. J.; Ning, Z.; Labelle, A. J.; Wei Chou, K.; Amassian, A.; Sargent, E. H. Hybrid Passivated Colloidal Quantum Dot Solids. Nat. Nanotechnol. 2012, 7, 577−582. (9) Shen, Q.; Kobayashi, J.; Diguna, L. J.; Toyoda, T. Effect of ZnS Coating on the Photovoltaic Properties of CdSe Quantum DotSensitized Solar Cells. J. Appl. Phys. 2008, 103, 084304−084309. (10) Hodes, G.; Albu-Yaron, A.; Decker, F.; Motisuke, P. ThreeDimensional Quantum-Size Effect in Chemically Deposited Cadmium Selenide Films. Phys. Rev. B 1987, 36, 4215−4221. (11) González-Pedro, V.; Xu, X.; Mora-Seró, I.; Bisquert, J. Modeling High-Efficiency Quantum Dot Sensitized Solar Cells. ACS Nano 2010, 4, 5783−5790. (12) Chen, J.; Zhao, D. W.; Song, J. L.; Sun, X. W.; Deng, X. W.; Liu, X. W.; Lei, W. Directly Assembled CdSe Quantum Dots on TiO2 in Aqueous Solution by Adjusting pH Value for Quantum Dot Sensitized Solar Cells. Electrochem. Commun. 2009, 11, 2265−2267. (13) Zhang, H.; Cheng, K.; Hou, Y. M.; Fang, Z.; Pan, Z. X.; Wu, W. J.; Hua, J. L.; Zhong, X. H. Efficient CdSe Quantum Dot-Sensitized Solar Cells Prepared by a Postsynthesis Assembly Approach. Chem. Commun. 2012, 48, 11235−11237. (14) Guijarro, N.; Shen, Q.; Giménez, S.; Mora-Seró, I.; Bisquert, J.; Lana-Villarreal, T.; Toyoda, T.; Gómez, R. Direct Correlation between Ultrafast Injection and Photoanode Performance in Quantum Dot Sensitized Solar Cells. J. Phys. Chem. C 2010, 114, 22352−22360. (15) Nevins, J.; Coughlin, K. M.; Watson, D. F. Attachment of CdSe Nanoparticles to TiO2 via Aqueous Linker-Assisted Assembly: Influence of Molecular Linkers on Electronic Properties and Interfacial Electron Transfer. ACS Appl. Mater. Interfaces 2011, 3, 4242−4253. (16) Mora-Seró, I.; Giménez, S.; Moehl, T.; Fabregat-Santiago, F.; Lana-Villareal, T.; Gómez, R.; Bisquert, J. Factors Determining the Photovoltaic Performance of a CdSe Quantum Dot Sensitized Solar Cell: the Role of the Linker Molecule and of the Counter Electrode. Nanotechnology 2008, 19, 424007. (17) Martsinovich, N.; Jones, D. R.; Troisi, A. Electronic Structure of TiO2 Surfaces and Effect of Molecular Adsorbates Using Different DFT Implementations. J. Phys. Chem. C 2010, 114, 22659−22670. (18) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (19) Delley, B. An All-Electron Numerical Method for Solving the Local Density Functional for Polyatomic Molecules. J. Chem. Phys. 1990, 92, 508−517.

Figure 4. Nyquist (top) and Bode (bottom) plots of Cys (black) and MPA (red) based QDSSCs.

occurs largely independently of the two linkers investigated. Furthermore, Cys cells were more effective in suppressing deleterious QDs aggregation. Within the context of the recent significant progress in QDSSC research, exemplified by the introduction of a new type of polysulfide electrolyte,29 Mndoping of quantum dots during in situ deposition,30 and the highly efficient linker based cell mentioned above,13 our results can contribute to a new QDSSC design paradigm.

■

ASSOCIATED CONTENT

* Supporting Information S

Experimental details, FTIR spectra of Cys on TiOx, additional AFM images, absorption spectra of photoelectrodes, SEM images. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (D.M.G.); clark@ chemie.uni-erlangen.de (T.C.). Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Deutsche Forschungsgemeinschaft as part of the Excellence Cluster “Engineering of Advanced Materials” and by the Bayerische Staatsregierung as part of the “Solar Technologies go Hybrid” initiative. Johannes Margraf is supported by a Beilstein Foundation Scholarship. 2437

dx.doi.org/10.1021/la3047609 | Langmuir 2013, 29, 2434−2438

Langmuir

Article

(20) Delley, B. From Molecules to Solids with the DMol3 Approach. J. Chem. Phys. 2000, 113, 7756−7764. (21) CRC Handbook of Chemistry and Physics, 85th ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 2005. (22) Kosmulski, M. The Significance of the Difference in the Point of Zero Charge between Rutile and Anatase. Adv. Colloid Interface Sci. 2002, 99, 255−264. (23) Sambur, J. B.; Riha, S. C.; Choi, D.; Parkinson, B. A. Influence of Surface Chemistry on the Binding and Electronic Coupling of CdSe Quantum Dots to Single Crystal TiO2 Surfaces. Langmuir 2010, 26, 4839−4847. (24) Ito, S.; Murakami, T. N.; Comte, P.; Liska, P.; Grätzel, C.; Nazeeruddin, M. K.; Grätzel, M. Fabrication of Thin Film Dye Sensitized Solar Cells with Solar to Electric Power Conversion Efficiency over 10%. Thin Solid Films 2008, 516, 4613−4619. (25) Guijarro, N.; Lana-Villarreal, T.; Mora-Seró, I.; Bisquert, J.; Gómez, R. CdSe Quantum Dot-Sensitized TiO2 Electrodes: Effect of Quantum Dot Coverage and Mode of Attachment. J. Phys. Chem. C 2009, 113, 4208−4214. (26) Zhu, G.; Xu, T.; Lv, T.; Pan, L.; Zhao, Q.; Sun, Z. GrapheneIncorporated Nanocrystalline TiO2 Films for CdS Quantum DotSensitized Solar Cells. J. Electroanal. Chem. 2010, 650, 248−251. (27) Wang, Q.; Moser, J.-E.; Grätzel, M. Electrochemical Impedance Spectroscopic Analysis of Dye-Sensitized Solar Cells. J. Phys. Chem. B 2005, 109, 14945−14953. (28) Pernik, D. R.; Tvrdy, K.; Radich, J. G.; Kamat, P. V. Tracking the Adsorption and Electron Injection Rates of CdSe Quantum Dots on TiO2: Linked versus Direct Attachment. J. Phys. Chem. C 2011, 115, 13511−13519. (29) Li, L.; Yang, X.; Gao, J.; Tian, H.; Zhao, J.; Hagfeldt, A.; Sun, L. Highly Efficient CdS Quantum Dot-Sensitized Solar Cells Based on a Modified Polysulfide Electrolyte. J. Am. Chem. Soc. 2011, 133, 8458− 8458. (30) Santra, P. K.; Kamat, P. V. Mn-Doped Quantum Dot Sensitized Solar Cells: A Strategy to Boost Efficiency over 5%. J. Am. Chem. Soc. 2012, 134, 2508−2511.

2438

dx.doi.org/10.1021/la3047609 | Langmuir 2013, 29, 2434−2438